- Email: [email protected]
Construction and Validation of the Addiction Susceptibility Questionnaire (ASQ)
Procedia - Social and Behavioral Sciences 00 (2011) 000–000 Procedia - Social and Behavioral Sciences 30 (2011) 1742 – 1747
Procedia Social and Behavioral Sciences www.elsevier.com/locate/procedia
WCPCG-2011
Construction and validation of the addiction susceptibility questionnaire (ASQ) Ali Zeinalia, Rogayeh Vahdatb a b
Islamic Azad University, Khoy Branch, Khoy, Iran Islamic Azad University, Urmia Branch, Urmia, Iran
Abstract The purpose of the present study was to construct and validate an Addiction Susceptibility Questionnaire (ASQ) as well as to present a discriminant equation to test the discriminant power (DP) of the ASQ to predict subjects’ level of Addiction Susceptibility (AS). Two hundred and forty participants from four distinct of Iran took part in this study 120 of whom were recovered addicts and 120 were not addicts at all. They were all selected by stratified random sampling method. Results showed that at least four factors underlie the construct of AS. There is a significant difference between addicts and healthy individuals in each of these factors. There is also a positive correlation between the factors and the discriminant function (DF). In fact, these factors are reliable predictors of the AS. The predictive power of the discriminant equation was 97.2%. The PSQ was found to be a highly acceptable measure of Addiction Susceptibility and it is appropriate for adolescents and adults in clinical, educational and research settings. Keywords: Addiction susceptibility; Construction; Validation; Questionnaire
1. Introduction Addiction is a complex and multifactor disease in which biological, psychological, social, and spiritual factors play a fundamental role (Galanter, 2006). Addiction can be better understood in a Multifactorial Developmental Process that emphasizes the factors leading to susceptibility during developmental periods (Zeinali, Vahdat, & Isavi, 2008). Addiction-Prone theory posits that certain individuals, as a result of unhealthy personalities, represent high risks for drug dependency if they are exposed to certain psychoactive drugs (Gendreau & Gendreau, 1970). In line with this theory, Availability-Proneness theory puts the emphasis on the interaction between level of psychosocial AS and possibility of drug availability. (Lettieri, Sayers, & Pearson, 1980). Barnes, Murray, Patton, Bentler, and Anderson (2000) in a series of studies mentioned the Addiction-Prone Personality (APP), and stated that having the characteristics of APP leads to alcohol and drug abuse. What are these characteristics? How do they develop and make individuals vulnerable to addiction? Studies have shown that pathological development and susceptibility for addiction play a fundamental role in developing addiction. Barnes, et al. (2000), proposed two environmental factors contributing significantly to developing APP characteristics: a) Low levels of parental care and b) High scores on parental APP characteristics. Lettieri, et al. (1980) stated that the results of different studies show that opium abusers suffer from a lot of psychological problems such as impulsivity, psychopathic or sociopathic traits, low tolerance of failure, weak egofunctions, depression, etc. before they develop addiction. Franke et al. (2003), referred to the development of
1877-0428 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the 2nd World Conference on Psychology, Counselling and Guidance. doi:10.1016/j.sbspro.2011.10.336
Ali Zeinali and / Procedia Social and Behavioral 30 000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/Vahdat Procedia – Social -and Behavioral SciencesSciences 00 (2011)
1743
* Corresponding author. Tel.: 09143409171; fax: 04413674272. E-mail address: [email protected].
susceptibility through the life span. Flagel, Vazquez, and Robinson (2003) discovered that early childhood events may change the neurobiological layers and help to develop a different kind of susceptibility for drug abuse in adolescents and adults. Dependence occurs only in a small number of individuals who take addictive drugs. This shows that there is a considerable difference in individuals’ level of susceptibility and drug dependence (Agatsuma & Hiroi, 2004). Hiroi and Agatsuma (2005) declared that the majority of substance users do not develop addiction to nicotine, alcohol, or opiates in response to chronic exposure to a drug. Despite the long-term use of drugs, some users do not develop addiction. On the other hand, some individuals are easily addicted to substances after only a brief exposure. Zeinali, et al. (2008), in an ex-post facto research, showed that different pathological background factors including personality traits, lifestyle, social and family relationships, beliefs and thoughts, feelings and emotions, and behaviors play an important role in the development of addiction. Addicts are significantly different from healthy individuals in all these pathological background factors even before they develop addiction. As mentioned, recent approaches and studies in different perspectives emphasize susceptibility for drug abuse. The present study is an attempt to introduce an instrument for assessing the ability of the instrument to measure AS. In this study, we will put emphasis on psychosocial susceptibility to addiction. Some scales have been constructed for measuring AS. Although these scales are efficient, they suffer from some drawbacks. Therefore, there seems to be a clear need for developing new scales. Some of the scales developed for measuring the AS construct are Addiction Potential Scale (APS) (Weed, Butcher, McKenna, & Ben-Porath, 1992), Addiction- Prone Personality or APP Test ( Barnes et al. , 2000 ), and Substance Abuse Subtle Screening Inventory (SASSI) ( Miller , 1985 ). The APS items were heterogeneous, and there seems to be no direct correlation between the items and drug abuse (Graham, 2006). Miller, shields, Campfield, Wallace, and Weiss (2007) in a meta-analysis showed that the APS scale has not proved to have acceptable reliability in most studies and clinical situations. The APP test, was designed for measuring individuals’ vulnerability to alcohol and drugs. This scale predicts vulnerability to alcohol and drugs. The APP was developed from the items of six scales (Baron's Ego-Strength scale, Eysenck Personality scale, Rosenberg's Self-Esteem scale, MacAndrew scale, Spielberger's Trait Anxiety scale, and Vando's Reducer-Augmenter scale) and considers AS in personality traits. Barnes, et al. (2000) argues that personality is only one of the factors predicting alcohol and drug abuse. Therefore, a comprehensive model is needed to take both personality and other factors into consideration. The SASSI, is designed to identify individuals who have a high probability of drug abuse and is used to screen adolescents with probability of substance abuse. But there is clearly a need for a scale that can measure the potential for developing addiction in adolescents and young adult before they even start taking drugs. In other words, it is necessary to identify and predict the level of AS in young people. The aim of present study is to construct and validate a questionnaire to measure subjects’ level of Addiction Susceptibility. 2. Method 2.1. Participants Data were collected in 2007 from recovered addicts (N = 1,240 – 95% male), who had abstained from drugs for more than one year. They were members of Narcotics Anonymous (N.A), Therapeutic Community (TC), and Group Therapy (GT). 240 participants were selected by stratified random sampling method on the basis of sex and city. According to the Rahimi et al (2000) study, Iranian provinces were divided in two parts: areas with high prevalence of addiction and areas with low prevalence of addiction. Out of the 32 provinces in Iran, Urmia and Ghom were selected randomly as areas with low prevalence, and Tehran and Shiraz as areas with high prevalence of addiction. Participants (5% female vs 95% male) were selected from provinces of these two parts. With respect to sex and city, the sample was 17.7% from Urmia, 12.9% from Gohm, 46.8% from Tehran, and 22.6% from Shiraz. 120 recovered addicts were taken as the experimental group and 120 healthy individuals were assigned to the control group from the same province. The control group was matched for age, education level and job with experimental group. 2.2. Materials and Procedure
1744
Ali Zeinali and Rogayeh Vahdat / Procedia - Social andSciences Behavioral Sciences 30 (2011) 1742 – 1747 Ali Zeinali / Procedia – Social and Behavioral 00 (2011) 000–000
In this study, we identified and classified important background factors that increase susceptibility for addiction under four categories (Behaviors, social relationships, personal traits and Beliefs); factors based on the addictionprone theory, factors indicated in studies related to developing AS (Hiroi and Agatsuma, 2005; Flagel at al., 2003); Agatsuma and Hiroi, 2004; Frank et al., 2003), factors derived from the information gathered during group treatment of addicts and their families, taking part in N.A and NAR-ANON free meetings, and factors inferred from studying the autobiographies received from 15 recovered addicts. Seventy items were designed. Three psychologists (PhD) evaluated these items and selected 50 items out of the 70 items. Next, we conducted an exploratory factor analysis on the items of the questionnaire and then estimated the validity and discriminant power (DP) of AS questionnaire. 3. Results In the first step, the items of questionnaire were analyzed using exploratory factor analysis with varimax rotation. All the 50 items of the questionnaire were loaded on the 4 factors and no item was omitted. Items indexes were optimal for becoming factors. KMO and Bartlett's Test of Sphericity were 0.961 and 10990.68 (Sig: 0.001) respectively. The four constructed factors were named according to the content of their items: Factor1- behaviors, 13 items. Factor2- lifestyle and social relationships, 14 items. Factor3- personality characters, 12 items. Factor4- emotions and beliefs, 11 items. In the second step, the reliability of the ASQ and its factors was estimated. The reliability of ASQ was .98 and its factors were.95, .97, .92 and .92 by Cronbach alpha method respectively. Table 1. Discriminant analysis with concurrent method for predicting variables
healthy
2.98
Std. Deviation 3.03
Addict
11.79
1.73
healthy
3.23
2.99
Addict
14.18
1.34
healthy
4.89
2.98
Addict
11.66
.59
healthy
3.47
2.34
Addict
9.73
.66
Variables* behaviors life style and social relationships Personality characteristics emotions and Beliefs
Mean
Wilks' Lambda
F
.320
710.90
.143
1267.30
.268
577.90
.217
763.76
*df1=1, df2=212, sig= 0.001
In the third step, the DP of the ASQ was tested (Table 1). The predictor (Independent) variables included behaviors, life style and social relationships, personality characters, and emotions and beliefs. The criterion (Dependent) variable was being addicted or non-addicted. The mean score for addicts in all the variables was high and very different from those of healthy individuals. High scores in any of those factors indicate the pathological status (high AS) of the individuals. Analysis of variance between addicts and non-addicts’ scores shows a significant difference between the two groups in all factors. Each of the factors as a predictor variable significantly affects the criterion variable. Furthermore, a low value of Wilks' Lambda for any of the predictor variables indicates a significant difference between the two groups. The table 2 shows, there is a function for analysis. High eigen value indicates that the function has a high DP. Its canonical correlation and its square show that 86% variance of the four predictor variables for the only DF is illustrated in the difference between the two groups. Considering the very low Lambda value, high Chi-square and its significance, we can say that the obtained DF has a good DP to distinguish addicts from healthy individuals. The centroid score for addicts is 2.36 and for healthy individuals it is -2.85. Obviously the function has the ability to distinguish between the two groups. Therefore, the cut-off point for being an addict or a non-addict in the DF is 0.49. If a person’s score in the DF tends (from the cut-off point) towards the centroid score for addicts, it is
1745
Ali Zeinali and / Procedia Social and Behavioral 30 000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/Vahdat Procedia – Social -and Behavioral SciencesSciences 00 (2011)
predicted that he/she will develop addiction susceptibility. If, on the other hand, his/her score tends towards the centroid score for healthy individuals, the prediction is that the person will maintain his/her health. Overall, the predictive power of the DF is 97.2%. The DF is able to distinguish correctly about 94.8% of healthy individuals and 99.1% of addicts. This indicates the high DP of the ASQ for distinguishing healthy individuals from addicts. Table 2 . Summary of Canonical discriminant Function Number of functions 1
Eigen value
Canonical Correlation
Eta square
Wilks' Lambda
Chisquare
df
sig
6.80
.93
.86
.13
431.51
4
.001
Centriod for addict 2.36
centroid for healthy individuals -2.58
Prediction of Group Membership 97.2%
The standardized and unstandardized coefficients of the discriminant Function was showen in table 3. There is a positive correlation between all the variables and the DF. Therefore, it can be concluded that the variables are reliable predictors of AS. Lifestyle and social relationships have the highest and personality characters the lowest correlation with the DF. Therefore, lifestyle and social relationships are of great value for distinguishing healthy individuals from addicts in the DF equation. If the score of subject is closer to the centroid for addicts, it is predicted that probably he/she will become addicted, and if the score is closer to the centroid for healthy individuals, the prediction is that he/she will probably stay healthy. Table 3 . Standardized and unstandardized coefficients of the discriminant Function variables
standardized coefficients
unstandardized coefficients
behaviors
.196
.082
life style and social relationships
.614
.274
personality characteristics
.135
.066
emotions and Beliefs
.276
.167
-
-4.833
constant
4. Discussion The items of ASQ were analyzed by explanatory factor analysis and 4 factors were determined as AS constructs. The reliability and validity of ASQ and its factors was evaluated and judged to be high. Four constructed factors were loaded for discriminant analysis (DA) in a predictive design by concurrent method. The results show a significant difference in all factors between addicts and healthy individuals’ scores. Each factor in predictor variables has a significant effect on the criterion variable. The DF can distinguish between addicts and healthy individuals which shows the high DP of ASQ for the detection of healthy individuals and addicts. There is a positive correlation between all the variables and the DF. In fact, the variables are good predictors of AS. The findings of the study show that the more an individual has had risky behaviors, pathological personality characters, unhealthy lifestyle, social relationships, and abnormal emotions and beliefs, the more he/she is likely to develop addiction susceptibility. This finding is in line with the results of other studies (Lettieri et al., 1980; Barnes et al., 2000; Agatsuma and Hiroi, 2004; Flagel et al., 2003; Hiroi and Agatsuma, 2005), and confirming their findings. The ASQ is regarded as a complement to other scales (APS, weed et al, 1992, APP, Barnes et al., 2000, SASSI, Miller, 1985) because of its factor structure, high reliability and good DP. More studies are needed to reveal ASQ’s strengths or weaknesses. Desire and interest in conducting an AS-oriented approach to psychological studies on addiction (Gendreau & Gendreau, 1970, Lettieri et al., 1980; Barnes et al., 2000) have faded during recent decades. The majority of researchers have an inclination towards conducting genetically-based studies and put emphasis on environmental factors (the role of learning and development) causing addiction (Agatsuma & Hiroi, 2004, Flagel et al., 2003, Hiroi
1746
Ali Zeinali and Rogayeh Vahdat / Procedia - Social andSciences Behavioral Sciences 30 (2011) 1742 – 1747 Ali Zeinali / Procedia – Social and Behavioral 00 (2011) 000–000
& Agatsuma, 2005). In this study, we will focus on psychosocial susceptibility for addiction. We feel the need to carry out more research studies from this perspective. Studies include unhealthy personality traits (Gendreau & Gendreau, 1970, Barnes et al., 2000), interaction of AS and drug availability ( Lettieri et al., 1980 ), early life events (Flagel et al.,2003), genetic variations and environmental factors ( Agatsuma & Hiroi , 2004; Hiroi & Agatsuma , 2005), as causes of vulnerability and susceptibility to addiction. The present study followed a psychosocial perspective. According to this perspective, unhealthy development in psychosocial aspects during childhood and adolescence will cause vulnerability and susceptibility for addiction. Addiction develops during developmental periods, parallel with the formation of risky behaviors, unhealthy lifestyle and social relationships, pathological personality characters and abnormal emotions and beliefs, even before the individual has abused the substance. In fact, addicts are individuals whose drug problem has been superimposed on an already unhealthy developmental background; it is not just drug abuse that causes addiction. In agreement with the studies in this field, the present study considers the issue from another perspective and emphasizes psychosocial susceptibility for becoming addicted and claims that AS can be measured by standardized instruments. The construction of AS is an important concept to the understanding of addiction development. Preventive programs should start in childhood and within families. Furthermore, in addiction treatment, not only the destroyed physical, psychological, social, and spiritual aspects but also the pathological susceptible background should be taken into consideration. Future studies can replicate this study with a larger sample and use other assessment methods such as the interview. Longitudinal studies should be carried out to broaden our understanding of AS. Further investigations can also be directed at studying the relationship between AS and other psychological constructs such as parenting style, attachment style, etc. And finally, the existence of AS among different groups of society such as high school students, university students, workers, etc. can be investigated. The ASQ is found to be a reliable measure of AS and it is appropriate for adolescents and adults in clinical, educational and research settings. References Agatsuma, S., & Hiroi, N. (2004). Genetic basis of drug dependence and comorbid behavioral trait. Japanese Journal of Psychopharmacology, 24(3),137-45. Barnes, G. E., Murray, R. P., Patton, D., Bentler, P. M., & Anderson, R.E. (2000). The addiction-prone personality. New york , Kluwer Academic/Plenum Publishers. Cooper, C. (2002). Individual differences (2nd ed.). New York: Oxford University Press. Flagel, S.B., Vázquez, D.M., & Robinson, T.E. (2003). Manipulations during the second week of life increase susceptibility to cocaine selfadministration in a stressor and gender-specific manner. Neuropsychopharmacology, 28, 1741-1751. Franke, P., Neef, D., Weiffenbach, Om., Gansicke, M., Hautzinger, M., Maier, W. (2003). Psychiatric comorbidity in risk groups of opioid addiction: a comparision between opioid dependent and nonopioid depedendent prisioners. Fortschr Nwurol Psychiatr, 71(1),37-44. Galanter, M. (2006). Innovations: alcohol and drug abuse: spirituality in alcoholics anonymous: a valuable adjunct to psychiatric services. Psychiatric Services, 57(3), 307-9. Gendreau, P., & Gendreau, L. P. (1970).The “addiction-prone” personality: A study of Canadian heroin addicts. Canadian Journal of Behavioral Science, 2, 18-25. Graham, J. R. (2006). MMPI-2: Assessing personality and psychopathology (4th edition). New York: Oxford University Press. Hiroi, N., & Agatsuma, S .(2005). Genetic susceptibility to substance dependence. Journal of Molecular Psychiatry, 10, 336–344. doi:10.1038/sj.mp.4001622 Lettieri, D. J., Sayers, M., & Pearson,H. W.(1980). Theories on drug abuse, selected contemporary perspectives. NIDA Research Monograph, U.S, Washington, D.C. Miller, G. A. (1985). The SASSI manual. Spencer, IN: Spencer Evening World. Miller, C. S., Shields, A. L., Campfield, D., Wallace, K. A., & Weiss, R. D.(2oo7). Substance use scales of the minnesota multiphasic personality inventory, An Exploration of score reliability via meta-analysis. Educational and Psychological Measurement, 67(6),1052-1065.
Ali Zeinali and Vahdat / Procedia - Social and Behavioral 30000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/ Procedia – Social and Behavioral SciencesSciences 00 (2011)
1747
doi:10.1177/0013164406299130 Rahimi, A; et al. (2000). Epidemiology of substance abuse in iran.Tehran: Iran University of medical science. Weed,N., Butcher,N.J., McKenna,T., & Ben-Porath,Y.(1992). New measures for assessing alcohol and other drug problems with MMPI-2: the APS & AAS. Journal of personality assessment, 58, 389-404. Zeinali, A., Vahdat, R., & Eisavi, M.(2008). Pre-addiction susceptibility backgrounds in recovered drug users. Iranian Journal of Psychiatry and Clinical Psychology, 14(1), 71- 79.
Procedia Social and Behavioral Sciences www.elsevier.com/locate/procedia
WCPCG-2011
Construction and validation of the addiction susceptibility questionnaire (ASQ) Ali Zeinalia, Rogayeh Vahdatb a b
Islamic Azad University, Khoy Branch, Khoy, Iran Islamic Azad University, Urmia Branch, Urmia, Iran
Abstract The purpose of the present study was to construct and validate an Addiction Susceptibility Questionnaire (ASQ) as well as to present a discriminant equation to test the discriminant power (DP) of the ASQ to predict subjects’ level of Addiction Susceptibility (AS). Two hundred and forty participants from four distinct of Iran took part in this study 120 of whom were recovered addicts and 120 were not addicts at all. They were all selected by stratified random sampling method. Results showed that at least four factors underlie the construct of AS. There is a significant difference between addicts and healthy individuals in each of these factors. There is also a positive correlation between the factors and the discriminant function (DF). In fact, these factors are reliable predictors of the AS. The predictive power of the discriminant equation was 97.2%. The PSQ was found to be a highly acceptable measure of Addiction Susceptibility and it is appropriate for adolescents and adults in clinical, educational and research settings. Keywords: Addiction susceptibility; Construction; Validation; Questionnaire
1. Introduction Addiction is a complex and multifactor disease in which biological, psychological, social, and spiritual factors play a fundamental role (Galanter, 2006). Addiction can be better understood in a Multifactorial Developmental Process that emphasizes the factors leading to susceptibility during developmental periods (Zeinali, Vahdat, & Isavi, 2008). Addiction-Prone theory posits that certain individuals, as a result of unhealthy personalities, represent high risks for drug dependency if they are exposed to certain psychoactive drugs (Gendreau & Gendreau, 1970). In line with this theory, Availability-Proneness theory puts the emphasis on the interaction between level of psychosocial AS and possibility of drug availability. (Lettieri, Sayers, & Pearson, 1980). Barnes, Murray, Patton, Bentler, and Anderson (2000) in a series of studies mentioned the Addiction-Prone Personality (APP), and stated that having the characteristics of APP leads to alcohol and drug abuse. What are these characteristics? How do they develop and make individuals vulnerable to addiction? Studies have shown that pathological development and susceptibility for addiction play a fundamental role in developing addiction. Barnes, et al. (2000), proposed two environmental factors contributing significantly to developing APP characteristics: a) Low levels of parental care and b) High scores on parental APP characteristics. Lettieri, et al. (1980) stated that the results of different studies show that opium abusers suffer from a lot of psychological problems such as impulsivity, psychopathic or sociopathic traits, low tolerance of failure, weak egofunctions, depression, etc. before they develop addiction. Franke et al. (2003), referred to the development of
1877-0428 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the 2nd World Conference on Psychology, Counselling and Guidance. doi:10.1016/j.sbspro.2011.10.336
Ali Zeinali and / Procedia Social and Behavioral 30 000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/Vahdat Procedia – Social -and Behavioral SciencesSciences 00 (2011)
1743
* Corresponding author. Tel.: 09143409171; fax: 04413674272. E-mail address: [email protected].
susceptibility through the life span. Flagel, Vazquez, and Robinson (2003) discovered that early childhood events may change the neurobiological layers and help to develop a different kind of susceptibility for drug abuse in adolescents and adults. Dependence occurs only in a small number of individuals who take addictive drugs. This shows that there is a considerable difference in individuals’ level of susceptibility and drug dependence (Agatsuma & Hiroi, 2004). Hiroi and Agatsuma (2005) declared that the majority of substance users do not develop addiction to nicotine, alcohol, or opiates in response to chronic exposure to a drug. Despite the long-term use of drugs, some users do not develop addiction. On the other hand, some individuals are easily addicted to substances after only a brief exposure. Zeinali, et al. (2008), in an ex-post facto research, showed that different pathological background factors including personality traits, lifestyle, social and family relationships, beliefs and thoughts, feelings and emotions, and behaviors play an important role in the development of addiction. Addicts are significantly different from healthy individuals in all these pathological background factors even before they develop addiction. As mentioned, recent approaches and studies in different perspectives emphasize susceptibility for drug abuse. The present study is an attempt to introduce an instrument for assessing the ability of the instrument to measure AS. In this study, we will put emphasis on psychosocial susceptibility to addiction. Some scales have been constructed for measuring AS. Although these scales are efficient, they suffer from some drawbacks. Therefore, there seems to be a clear need for developing new scales. Some of the scales developed for measuring the AS construct are Addiction Potential Scale (APS) (Weed, Butcher, McKenna, & Ben-Porath, 1992), Addiction- Prone Personality or APP Test ( Barnes et al. , 2000 ), and Substance Abuse Subtle Screening Inventory (SASSI) ( Miller , 1985 ). The APS items were heterogeneous, and there seems to be no direct correlation between the items and drug abuse (Graham, 2006). Miller, shields, Campfield, Wallace, and Weiss (2007) in a meta-analysis showed that the APS scale has not proved to have acceptable reliability in most studies and clinical situations. The APP test, was designed for measuring individuals’ vulnerability to alcohol and drugs. This scale predicts vulnerability to alcohol and drugs. The APP was developed from the items of six scales (Baron's Ego-Strength scale, Eysenck Personality scale, Rosenberg's Self-Esteem scale, MacAndrew scale, Spielberger's Trait Anxiety scale, and Vando's Reducer-Augmenter scale) and considers AS in personality traits. Barnes, et al. (2000) argues that personality is only one of the factors predicting alcohol and drug abuse. Therefore, a comprehensive model is needed to take both personality and other factors into consideration. The SASSI, is designed to identify individuals who have a high probability of drug abuse and is used to screen adolescents with probability of substance abuse. But there is clearly a need for a scale that can measure the potential for developing addiction in adolescents and young adult before they even start taking drugs. In other words, it is necessary to identify and predict the level of AS in young people. The aim of present study is to construct and validate a questionnaire to measure subjects’ level of Addiction Susceptibility. 2. Method 2.1. Participants Data were collected in 2007 from recovered addicts (N = 1,240 – 95% male), who had abstained from drugs for more than one year. They were members of Narcotics Anonymous (N.A), Therapeutic Community (TC), and Group Therapy (GT). 240 participants were selected by stratified random sampling method on the basis of sex and city. According to the Rahimi et al (2000) study, Iranian provinces were divided in two parts: areas with high prevalence of addiction and areas with low prevalence of addiction. Out of the 32 provinces in Iran, Urmia and Ghom were selected randomly as areas with low prevalence, and Tehran and Shiraz as areas with high prevalence of addiction. Participants (5% female vs 95% male) were selected from provinces of these two parts. With respect to sex and city, the sample was 17.7% from Urmia, 12.9% from Gohm, 46.8% from Tehran, and 22.6% from Shiraz. 120 recovered addicts were taken as the experimental group and 120 healthy individuals were assigned to the control group from the same province. The control group was matched for age, education level and job with experimental group. 2.2. Materials and Procedure
1744
Ali Zeinali and Rogayeh Vahdat / Procedia - Social andSciences Behavioral Sciences 30 (2011) 1742 – 1747 Ali Zeinali / Procedia – Social and Behavioral 00 (2011) 000–000
In this study, we identified and classified important background factors that increase susceptibility for addiction under four categories (Behaviors, social relationships, personal traits and Beliefs); factors based on the addictionprone theory, factors indicated in studies related to developing AS (Hiroi and Agatsuma, 2005; Flagel at al., 2003); Agatsuma and Hiroi, 2004; Frank et al., 2003), factors derived from the information gathered during group treatment of addicts and their families, taking part in N.A and NAR-ANON free meetings, and factors inferred from studying the autobiographies received from 15 recovered addicts. Seventy items were designed. Three psychologists (PhD) evaluated these items and selected 50 items out of the 70 items. Next, we conducted an exploratory factor analysis on the items of the questionnaire and then estimated the validity and discriminant power (DP) of AS questionnaire. 3. Results In the first step, the items of questionnaire were analyzed using exploratory factor analysis with varimax rotation. All the 50 items of the questionnaire were loaded on the 4 factors and no item was omitted. Items indexes were optimal for becoming factors. KMO and Bartlett's Test of Sphericity were 0.961 and 10990.68 (Sig: 0.001) respectively. The four constructed factors were named according to the content of their items: Factor1- behaviors, 13 items. Factor2- lifestyle and social relationships, 14 items. Factor3- personality characters, 12 items. Factor4- emotions and beliefs, 11 items. In the second step, the reliability of the ASQ and its factors was estimated. The reliability of ASQ was .98 and its factors were.95, .97, .92 and .92 by Cronbach alpha method respectively. Table 1. Discriminant analysis with concurrent method for predicting variables
healthy
2.98
Std. Deviation 3.03
Addict
11.79
1.73
healthy
3.23
2.99
Addict
14.18
1.34
healthy
4.89
2.98
Addict
11.66
.59
healthy
3.47
2.34
Addict
9.73
.66
Variables* behaviors life style and social relationships Personality characteristics emotions and Beliefs
Mean
Wilks' Lambda
F
.320
710.90
.143
1267.30
.268
577.90
.217
763.76
*df1=1, df2=212, sig= 0.001
In the third step, the DP of the ASQ was tested (Table 1). The predictor (Independent) variables included behaviors, life style and social relationships, personality characters, and emotions and beliefs. The criterion (Dependent) variable was being addicted or non-addicted. The mean score for addicts in all the variables was high and very different from those of healthy individuals. High scores in any of those factors indicate the pathological status (high AS) of the individuals. Analysis of variance between addicts and non-addicts’ scores shows a significant difference between the two groups in all factors. Each of the factors as a predictor variable significantly affects the criterion variable. Furthermore, a low value of Wilks' Lambda for any of the predictor variables indicates a significant difference between the two groups. The table 2 shows, there is a function for analysis. High eigen value indicates that the function has a high DP. Its canonical correlation and its square show that 86% variance of the four predictor variables for the only DF is illustrated in the difference between the two groups. Considering the very low Lambda value, high Chi-square and its significance, we can say that the obtained DF has a good DP to distinguish addicts from healthy individuals. The centroid score for addicts is 2.36 and for healthy individuals it is -2.85. Obviously the function has the ability to distinguish between the two groups. Therefore, the cut-off point for being an addict or a non-addict in the DF is 0.49. If a person’s score in the DF tends (from the cut-off point) towards the centroid score for addicts, it is
1745
Ali Zeinali and / Procedia Social and Behavioral 30 000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/Vahdat Procedia – Social -and Behavioral SciencesSciences 00 (2011)
predicted that he/she will develop addiction susceptibility. If, on the other hand, his/her score tends towards the centroid score for healthy individuals, the prediction is that the person will maintain his/her health. Overall, the predictive power of the DF is 97.2%. The DF is able to distinguish correctly about 94.8% of healthy individuals and 99.1% of addicts. This indicates the high DP of the ASQ for distinguishing healthy individuals from addicts. Table 2 . Summary of Canonical discriminant Function Number of functions 1
Eigen value
Canonical Correlation
Eta square
Wilks' Lambda
Chisquare
df
sig
6.80
.93
.86
.13
431.51
4
.001
Centriod for addict 2.36
centroid for healthy individuals -2.58
Prediction of Group Membership 97.2%
The standardized and unstandardized coefficients of the discriminant Function was showen in table 3. There is a positive correlation between all the variables and the DF. Therefore, it can be concluded that the variables are reliable predictors of AS. Lifestyle and social relationships have the highest and personality characters the lowest correlation with the DF. Therefore, lifestyle and social relationships are of great value for distinguishing healthy individuals from addicts in the DF equation. If the score of subject is closer to the centroid for addicts, it is predicted that probably he/she will become addicted, and if the score is closer to the centroid for healthy individuals, the prediction is that he/she will probably stay healthy. Table 3 . Standardized and unstandardized coefficients of the discriminant Function variables
standardized coefficients
unstandardized coefficients
behaviors
.196
.082
life style and social relationships
.614
.274
personality characteristics
.135
.066
emotions and Beliefs
.276
.167
-
-4.833
constant
4. Discussion The items of ASQ were analyzed by explanatory factor analysis and 4 factors were determined as AS constructs. The reliability and validity of ASQ and its factors was evaluated and judged to be high. Four constructed factors were loaded for discriminant analysis (DA) in a predictive design by concurrent method. The results show a significant difference in all factors between addicts and healthy individuals’ scores. Each factor in predictor variables has a significant effect on the criterion variable. The DF can distinguish between addicts and healthy individuals which shows the high DP of ASQ for the detection of healthy individuals and addicts. There is a positive correlation between all the variables and the DF. In fact, the variables are good predictors of AS. The findings of the study show that the more an individual has had risky behaviors, pathological personality characters, unhealthy lifestyle, social relationships, and abnormal emotions and beliefs, the more he/she is likely to develop addiction susceptibility. This finding is in line with the results of other studies (Lettieri et al., 1980; Barnes et al., 2000; Agatsuma and Hiroi, 2004; Flagel et al., 2003; Hiroi and Agatsuma, 2005), and confirming their findings. The ASQ is regarded as a complement to other scales (APS, weed et al, 1992, APP, Barnes et al., 2000, SASSI, Miller, 1985) because of its factor structure, high reliability and good DP. More studies are needed to reveal ASQ’s strengths or weaknesses. Desire and interest in conducting an AS-oriented approach to psychological studies on addiction (Gendreau & Gendreau, 1970, Lettieri et al., 1980; Barnes et al., 2000) have faded during recent decades. The majority of researchers have an inclination towards conducting genetically-based studies and put emphasis on environmental factors (the role of learning and development) causing addiction (Agatsuma & Hiroi, 2004, Flagel et al., 2003, Hiroi
1746
Ali Zeinali and Rogayeh Vahdat / Procedia - Social andSciences Behavioral Sciences 30 (2011) 1742 – 1747 Ali Zeinali / Procedia – Social and Behavioral 00 (2011) 000–000
& Agatsuma, 2005). In this study, we will focus on psychosocial susceptibility for addiction. We feel the need to carry out more research studies from this perspective. Studies include unhealthy personality traits (Gendreau & Gendreau, 1970, Barnes et al., 2000), interaction of AS and drug availability ( Lettieri et al., 1980 ), early life events (Flagel et al.,2003), genetic variations and environmental factors ( Agatsuma & Hiroi , 2004; Hiroi & Agatsuma , 2005), as causes of vulnerability and susceptibility to addiction. The present study followed a psychosocial perspective. According to this perspective, unhealthy development in psychosocial aspects during childhood and adolescence will cause vulnerability and susceptibility for addiction. Addiction develops during developmental periods, parallel with the formation of risky behaviors, unhealthy lifestyle and social relationships, pathological personality characters and abnormal emotions and beliefs, even before the individual has abused the substance. In fact, addicts are individuals whose drug problem has been superimposed on an already unhealthy developmental background; it is not just drug abuse that causes addiction. In agreement with the studies in this field, the present study considers the issue from another perspective and emphasizes psychosocial susceptibility for becoming addicted and claims that AS can be measured by standardized instruments. The construction of AS is an important concept to the understanding of addiction development. Preventive programs should start in childhood and within families. Furthermore, in addiction treatment, not only the destroyed physical, psychological, social, and spiritual aspects but also the pathological susceptible background should be taken into consideration. Future studies can replicate this study with a larger sample and use other assessment methods such as the interview. Longitudinal studies should be carried out to broaden our understanding of AS. Further investigations can also be directed at studying the relationship between AS and other psychological constructs such as parenting style, attachment style, etc. And finally, the existence of AS among different groups of society such as high school students, university students, workers, etc. can be investigated. The ASQ is found to be a reliable measure of AS and it is appropriate for adolescents and adults in clinical, educational and research settings. References Agatsuma, S., & Hiroi, N. (2004). Genetic basis of drug dependence and comorbid behavioral trait. Japanese Journal of Psychopharmacology, 24(3),137-45. Barnes, G. E., Murray, R. P., Patton, D., Bentler, P. M., & Anderson, R.E. (2000). The addiction-prone personality. New york , Kluwer Academic/Plenum Publishers. Cooper, C. (2002). Individual differences (2nd ed.). New York: Oxford University Press. Flagel, S.B., Vázquez, D.M., & Robinson, T.E. (2003). Manipulations during the second week of life increase susceptibility to cocaine selfadministration in a stressor and gender-specific manner. Neuropsychopharmacology, 28, 1741-1751. Franke, P., Neef, D., Weiffenbach, Om., Gansicke, M., Hautzinger, M., Maier, W. (2003). Psychiatric comorbidity in risk groups of opioid addiction: a comparision between opioid dependent and nonopioid depedendent prisioners. Fortschr Nwurol Psychiatr, 71(1),37-44. Galanter, M. (2006). Innovations: alcohol and drug abuse: spirituality in alcoholics anonymous: a valuable adjunct to psychiatric services. Psychiatric Services, 57(3), 307-9. Gendreau, P., & Gendreau, L. P. (1970).The “addiction-prone” personality: A study of Canadian heroin addicts. Canadian Journal of Behavioral Science, 2, 18-25. Graham, J. R. (2006). MMPI-2: Assessing personality and psychopathology (4th edition). New York: Oxford University Press. Hiroi, N., & Agatsuma, S .(2005). Genetic susceptibility to substance dependence. Journal of Molecular Psychiatry, 10, 336–344. doi:10.1038/sj.mp.4001622 Lettieri, D. J., Sayers, M., & Pearson,H. W.(1980). Theories on drug abuse, selected contemporary perspectives. NIDA Research Monograph, U.S, Washington, D.C. Miller, G. A. (1985). The SASSI manual. Spencer, IN: Spencer Evening World. Miller, C. S., Shields, A. L., Campfield, D., Wallace, K. A., & Weiss, R. D.(2oo7). Substance use scales of the minnesota multiphasic personality inventory, An Exploration of score reliability via meta-analysis. Educational and Psychological Measurement, 67(6),1052-1065.
Ali Zeinali and Vahdat / Procedia - Social and Behavioral 30000–000 (2011) 1742 – 1747 AliRogayeh Zeinali/ Procedia – Social and Behavioral SciencesSciences 00 (2011)
1747
doi:10.1177/0013164406299130 Rahimi, A; et al. (2000). Epidemiology of substance abuse in iran.Tehran: Iran University of medical science. Weed,N., Butcher,N.J., McKenna,T., & Ben-Porath,Y.(1992). New measures for assessing alcohol and other drug problems with MMPI-2: the APS & AAS. Journal of personality assessment, 58, 389-404. Zeinali, A., Vahdat, R., & Eisavi, M.(2008). Pre-addiction susceptibility backgrounds in recovered drug users. Iranian Journal of Psychiatry and Clinical Psychology, 14(1), 71- 79.